Speech signals include information about the creator of the speech; that is, the person talking. Modern technology known as Speaker ID or Voice Verification can identify the speaker as one of a moderate collection of speakers, or can verify a claimed identity, in a short segment of speech or a collection of such segments. The requirement for such an identification process is that the speech of the person to be identified is available independently of the message.
The term voice recognition or speaker identification refers to finding the identity of “who” is speaking, rather than identifying the content of what they are saying. Recognizing the speaker can simplify tasks of translating speech in systems that have been trained on a specific person's voice or it can be used to authenticate or verify the identity of a speaker as part of a security process.
Biometrics refers to metrics related to human characteristics and traits. Biometric identification (or biometric authentication) is used in computer science as a form of identification and access control based on human characteristics and traits. It is also used to identify individuals in groups that may be monitored or under surveillance. Biometric identifiers are the distinctive, measurable characteristics used to label and describe individuals. Biometric identifiers are often categorized as physiological versus behavioral characteristics.